import pandas as pd
import numpy as np
from itertools import *

Step 0: Business context

Business goals: Answer the following questions

1. what is the leading smartphone company
2. what model is the leading model in the word
3. what model has the best margins

-> data only has a few features. Answer these questions instead:

1. what is the most popular smartphone company, by rating
2. what is the most popular model of phone, by rating
3. what aspects of the models are popular? -> categorized as such:

• I) brand questions
  • what is the most popular smartphone brand?
    • "Top Tiers"
    • "Consistency"
    • "Niche brands"
• II) model questions
  • what are the most popular models amongst all brands?
• III) specs questions
  • price: what is the most expensive model?
  • speed: what is the fastest model?
  • camera: what model has the best camera?
  • resolution: what model has the best resolution?
  • battery: what model has the biggest battery?

Step 1: load in data

pd.set_option('display.max_rows', None)
pd.set_option('display.max_columns', None)
pd.set_option('mode.chained_assignment',None)
np.set_printoptions(threshold=np.inf)

def open_data(repo, file):
    with open("".join([repo, file]), "r", errors="ignore") as r:
        file = pd.read_csv(r, engine="python")
        print(file.info())
    return file

repo = r"C:/Users/benno/OneDrive/Python/Dash/Deployment_ready/smartphones/assets/data/"
file = "smartphones.csv"

smart = open_data(repo, file)

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 980 entries, 0 to 979
Data columns (total 22 columns):
 #   Column                     Non-Null Count  Dtype  
---  ------                     --------------  -----  
 0   brand_name              980 non-null    object 
 1   model                      980 non-null    object 
 2   price                      980 non-null    int64  
 3   avg_rating                 879 non-null    float64
 4   5G_or_not                  980 non-null    int64  
 5   processor_brand            960 non-null    object 
 6   num_cores                  974 non-null    float64
 7   processor_speed            938 non-null    float64
 8   battery_capacity           969 non-null    float64
 9   fast_charging_available    980 non-null    int64  
 10  fast_charging              769 non-null    float64
 11  ram_capacity               980 non-null    int64  
 12  internal_memory            980 non-null    int64  
 13  screen_size                980 non-null    float64
 14  refresh_rate               980 non-null    int64  
 15  num_rear_cameras           980 non-null    int64  
 16  os                         966 non-null    object 
 17  primary_camera_rear        980 non-null    float64
 18  primary_camera_front       975 non-null    float64
 19  extended_memory_available  980 non-null    int64  
 20  resolution_height          980 non-null    int64  
 21  resolution_width           980 non-null    int64  
dtypes: float64(8), int64(10), object(4)
memory usage: 168.6+ KB
None

Step 2: Summaries

import warnings
warnings.filterwarnings("ignore")

def summary_dataframe_small(df):

	return pd.DataFrame({'features': df.columns,'dtypes': [df.dtypes[i] for i in range(len(df.columns))],
	'nulls':[df[df.columns[i]].isnull().sum() for i in range(len(df.columns))],'nonnuls / 1.0': (df.shape[0] - 
	df.isnull().sum().values) / df.shape[0],'uniques': [df[j].nunique() for j in df.columns]})

summary_dataframe_small(smart)

	features	dtypes	nulls	nonnuls / 1.0	uniques
0	brand_name	object	0	1.000000	46
1	model	object	0	1.000000	980
2	price	int64	0	1.000000	379
3	avg_rating	float64	101	0.896939	30
4	5G_or_not	int64	0	1.000000	2
5	processor_brand	object	20	0.979592	13
6	num_cores	float64	6	0.993878	3
7	processor_speed	float64	42	0.957143	35
8	battery_capacity	float64	11	0.988776	89
9	fast_charging_available	int64	0	1.000000	2
10	fast_charging	float64	211	0.784694	32
11	ram_capacity	int64	0	1.000000	9
12	internal_memory	int64	0	1.000000	8
13	screen_size	float64	0	1.000000	79
14	refresh_rate	int64	0	1.000000	6
15	num_rear_cameras	int64	0	1.000000	4
16	os	object	14	0.985714	3
17	primary_camera_rear	float64	0	1.000000	18
18	primary_camera_front	float64	5	0.994898	19
19	extended_memory_available	int64	0	1.000000	2
20	resolution_height	int64	0	1.000000	65
21	resolution_width	int64	0	1.000000	40

def summary_dataframe_large(df):
    
    summary = pd.DataFrame({
        "feature": df.columns,
        "dtype" : [str(df[col].dtype) for col in df.columns],
        "nulls": [f"{df[col].isnull().sum()}" for col in df.columns],
        "nonnulls%": [f"{round(((1 - df[col].isnull().sum() / len(df)) * 100), 2)}%" for col in df.columns],
        "STATS": ["|" for col in df.columns],
        "min": [df[col].min().round(2) if any(t in str(df[col].dtype) for t in ("float", "int")) else " " for col in df.columns],
        "50%": [df[col].median().round(2) if any(t in str(df[col].dtype) for t in ("float", "int")) else " "for col in df.columns],
        "max": [df[col].max().round(2) if any(t in str(df[col].dtype) for t in ("float", "int")) else " " for col in df.columns],
        "stdev": [df[col].std().round(2) if any(t in str(df[col].dtype) for t in ("float", "int")) else " " for col in df.columns],
        "mean": [df[col].mean().round(2) if any(t in str(df[col].dtype) for t in ("float", "int")) else " " for col in df.columns],
        "skew": ["" if not any(t in str(df[col].dtype) for t in ("float", "int")) else "left" if df[col].median() > df[col].mean() else "right" if df[col].median() < df[col].mean() else "equal" for col in df.columns],
        "cat_mode": [" " if any(t in str(df[col].dtype) for t in ("float", "int")) else df[col].mode()[0] for col in df.columns],        
        "cat_mode": [" " if any(t in str(df[col].dtype) for t in ("float", "int")) else df[col].mode()[0] for col in df.columns],
        "COUNTS": ["|" for col in df.columns],
        "nuniques": [df[col].nunique() for col in df.columns],
        "vcs-": [df[col].value_counts().reset_index()["count"].min() for col in df.columns],
        "vcs+": [df[col].value_counts().reset_index()["count"].max() for col in df.columns],
        "RANDOM": ["|" for col in df.columns],
        "row[random]" :[df[col].iloc[np.random.randint(len(df))].round(2) if any(t in str(df[col].dtype) for t in ("float", "int")) else df[col].iloc[np.random.randint(len(df))] for col in df.columns],
    })

    print(df.shape)

    return summary
    
summary = summary_dataframe_large(smart)
summary

(980, 22)

	feature	dtype	nulls	nonnulls%	STATS	min	50%	max	stdev	mean	skew	cat_mode	COUNTS	nuniques	vcs-	vcs+	RANDOM	row[random]
0	brand_name	object	0	100.0%	|							xiaomi	|	46	1	134	|	vivo
1	model	object	0	100.0%	|							Apple iPhone 11	|	980	1	1	|	Vivo Y20G
2	price	int64	0	100.0%	|	3499	19994.5	650000	39531.81	32520.5	right		|	379	1	21	|	12999
3	avg_rating	float64	101	89.69%	|	6.0	8.0	8.9	0.74	7.83	left		|	30	10	60	|	7.8
4	5G_or_not	int64	0	100.0%	|	0	1.0	1	0.5	0.56	left		|	2	431	549	|	1
5	processor_brand	object	20	97.96%	|							snapdragon	|	13	1	413	|	dimensity
6	num_cores	float64	6	99.39%	|	4.0	8.0	8.0	0.84	7.77	left		|	3	36	899	|	8.0
7	processor_speed	float64	42	95.71%	|	1.2	2.3	3.22	0.46	2.43	right		|	35	1	146	|	2.3
8	battery_capacity	float64	11	98.88%	|	1821.0	5000.0	22000.0	1009.54	4817.75	left		|	89	1	487	|	5000.0
9	fast_charging_available	int64	0	100.0%	|	0	1.0	1	0.35	0.85	left		|	2	143	837	|	1
10	fast_charging	float64	211	78.47%	|	10.0	33.0	240.0	34.28	46.13	right		|	32	1	152	|	65.0
11	ram_capacity	int64	0	100.0%	|	1	6.0	18	2.74	6.56	right		|	9	2	339	|	4
12	internal_memory	int64	0	100.0%	|	8	128.0	1024	107.13	141.04	right		|	8	1	523	|	128
13	screen_size	float64	0	100.0%	|	3.54	6.58	8.03	0.35	6.54	left		|	79	1	119	|	6.0
14	refresh_rate	int64	0	100.0%	|	60	90.0	240	28.99	92.26	right		|	6	1	368	|	60
15	num_rear_cameras	int64	0	100.0%	|	1	3.0	4	0.78	2.81	left		|	4	65	551	|	4
16	os	object	14	98.57%	|							android	|	3	11	909	|	android
17	primary_camera_rear	float64	0	100.0%	|	2.0	50.0	200.0	33.0	50.32	right		|	18	1	333	|	64.0
18	primary_camera_front	float64	5	99.49%	|	0.0	16.0	60.0	10.88	16.59	right		|	19	1	307	|	8.0
19	extended_memory_available	int64	0	100.0%	|	0	1.0	1	0.48	0.63	left		|	2	362	618	|	1
20	resolution_height	int64	0	100.0%	|	480	2400.0	3840	516.48	2214.66	left		|	65	1	344	|	1600
21	resolution_width	int64	0	100.0%	|	480	1080.0	2460	290.16	1075.85	left		|	40	1	593	|	1080

smart.columns

Index(['brand_name', 'model', 'price', 'avg_rating', '5G_or_not',
       'processor_brand', 'num_cores', 'processor_speed', 'battery_capacity',
       'fast_charging_available', 'fast_charging', 'ram_capacity',
       'internal_memory', 'screen_size', 'refresh_rate', 'num_rear_cameras',
       'os', 'primary_camera_rear', 'primary_camera_front',
       'extended_memory_available', 'resolution_height', 'resolution_width'],
      dtype='object')

smart.columns = ["brand", 'model', 'price', 'avg_rating', '5G_or_not',
       'processor_brand', 'num_cores', 'processor_speed', 'battery_capacity',
       'fast_charging_available', 'fast_charging', 'ram_capacity',
       'internal_memory', 'screen_size', 'refresh_rate', 'num_rear_cameras',
       'os', 'primary_camera_rear', 'primary_camera_front',
       'extended_memory_available', 'resolution_height', 'resolution_width']

# interested in the following columns
cols = [
    "brand", 
    "model", 
    "price", 
    "avg_rating", 
    "processor_speed", 
    "battery_capacity", 
    "primary_camera_rear", 
    "resolution_height", 
    "resolution_width"
]

smarts = smart[cols]

Step 3: Feature Engineering

smarts["total_res"] = smarts["resolution_height"] * smarts["resolution_width"]

smarts = smarts.drop(["resolution_height", "resolution_width"], axis=1)

Step 4: EDA

# truncated for final version

import plotly.io as pio
from plotly.offline import *
import plotly.express as px
import plotly.graph_objs as go
from plotly.subplots import make_subplots

Summary of data

• Indian market
• scores range from 6.1-8.9
• hundreds of phone models
• dozens of phone brands
• many brands have a range of scores
• features are:
  • Brand
  • Model
  • Price
  • Avg rating, 1-10
  • Processor Speed
  • Battery Capcity (size)
  • Primary camera
  • Total resolution
• 980 observations
  • total shape (980, 8)
  • total size 7840
  • mostly categorical ordinal, some continuous

smarts.size

7840

gb4 = smarts.groupby(["model","brand"]).agg({"avg_rating": "mean"}).reset_index()

px.strip(gb4.dropna(), x="model", y="avg_rating").update_layout({"height": 650})

Step 6: What are the most popular smartphones?

1. This is a ticky question. By pure mean, it's a company called Leitz. This is a German camera company, who only released one phone, and it was top-rated. Lack of sample size means the rating for the brand has stayed at that rating.

2. This analysis defines "top brand" in three ways:

• A) "Feeding frenzy" measurement: a brand's best phone (max rating, all models)
• B) "Consistency" brands: a brand's total average rating, regardless of how many phones
• C) "Niche" brands: a brand's total average rating, with population of brands with >= 10 phones
• D) "Top Tiers": brands with the most quantity of top-rated phones

D: "Top-Tier", most amount of maxed out phones

arrange data by brand

g = px.strip(gb4.dropna(), x="model", y="avg_rating", color="brand", color_discrete_sequence=px.colors.qualitative.T10).update_layout({"height": 650})
g

zoom in on phones with the maxium rating

g.add_shape(type="rect", 
              x0=0, y0=8.7, x1=900, y1=9.1,
              line=dict(color="#6d84c8", width=2))

many brands have top-rated phones (max rating = 8.9)

hr_mk = ( gb4["avg_rating"] == 8.9 )

h = px.scatter(gb4[hr_mk], x="model", y="avg_rating", size="avg_rating", color="brand", color_discrete_sequence=px.colors.qualitative.T10)
h

h.add_shape(
    type="rect",
    x0=7.5, x1=12.5,
    y0=8.5, y1=9.25,
    line={"color": "#6d84c8"},
)
h.add_shape(
    type="rect",
    x0=18.5, x1=25.5,
    y0=8.5, y1=9.25,
    line={"color": "#6d84c8"},
)
h.add_shape(
    type="rect",
    x0=28.5, x1=33.5,
    y0=8.5, y1=9.25,
    line={"color": "#6d84c8"},
)

Answer, 1D)

in the absence of sales data and with the inclusion of outlier brands who produce just 1 phone*, most popular smartphone brand Top Tier brands are:

• Samsung, with 7 top-rated phones.
• Motorola and Xiaomi come in at second
• Oneplus comes in at third

gb4[hr_mk].groupby(["brand"]).count().reset_index().sort_values(by="avg_rating", ascending=False).reset_index(drop=True)

	brand	model	avg_rating
0	samsung	7	7
1	motorola	5	5
2	xiaomi	5	5
3	oneplus	3	3
4	realme	2	2
5	iqoo	2	2
6	sony	2	2
7	huawei	1	1
8	lg	1	1
9	lenovo	1	1
10	leitz	1	1
11	infinix	1	1
12	google	1	1
13	oppo	1	1
14	tecno	1	1
15	zte	1	1

Part 1C)

Data ascending by pure mean:

• Average of all phones

gb1 = smarts.groupby(["brand"]).agg({"avg_rating" :"mean"}).reset_index()
gb1 = gb1.dropna().sort_values(by="avg_rating", ascending=True).reset_index(drop=True)
g = px.bar(gb1.dropna(), y="brand", x="avg_rating", color="avg_rating")

g.update_layout({"height": 400, "width": 450, 
                 "yaxis": {"tickangle": 45, "tickfont": {"size": 8}}})

let's add in "count"

gb2.sort_values(by="avg_rating_mean", ascending=False).reset_index(drop=True)[:10]

We can think of these brands as an "under the radar" cohort of smartphone brands.

Insight: Why is this important?

1. less homogeneity of brands creates an opportunity to for consumers to "be different" with their choice of smartphone brand
   • example: Nothing phone, "minimalist"
2. some smartphones may be ultra-specific, serving only one market niche
   • example: Doogee, battery life

gb2 = smarts.groupby(["brand"]).agg({"avg_rating": ["count", "mean"]}).reset_index()
gb2.columns = ["brand", "count", "avg_rating_mean"]
gb2 = gb2.sort_values(by="avg_rating_mean", ascending=True)

gb2.sort_values(by="avg_rating_mean", ascending=False).reset_index(drop=True)[:10]

	brand	count	avg_rating_mean
0	leitz	1	8.90
1	lenovo	2	8.80
2	sharp	1	8.80
3	asus	6	8.70
4	lg	2	8.70
5	royole	2	8.70
6	doogee	2	8.60
7	zte	2	8.55
8	nothing	5	8.52
9	blu	1	8.50

g = px.bar(gb2.dropna(), y="brand", x="avg_rating_mean", color="count", color_continuous_scale="plasma")

g.update_layout({"height": 400, "width": 450, 
                 "yaxis": {"tickangle": 45, "tickfont": {"size": 8}}})

ct_mk = ( gb2["count"] >= 10) 
g = px.bar(gb2[ct_mk].dropna(), y="brand", x="avg_rating_mean", color="avg_rating_mean", color_continuous_scale=[px.colors.sequential.Plasma[-1]] * 100)

g = px.scatter(gb2, x="avg_rating_mean", y="avg_rating_mean", hover_data="brand", size="count", color="count").update_traces({"marker":{"line": {"width": 1, "color": "black"}}}).update_layout({"width": 450, "height": 400})
g.add_annotation(x=8.7, y=8.9,
            text="Niche brands",
            showarrow=True,
            arrowhead=1)

Organized into a barchart. Brands with < 10 phones.

• Remember that the "average problem" still applies. (top rated brand has only produced one phone, and it just happens to by top-rated. Doesn't mean that brand is the top smartphone maker in the market.)

gb13 = smarts.groupby(["brand"]).agg({"avg_rating": ["count", "mean"]}).reset_index()
gb13.columns = ["brand", "count", "mean"]

lc_mk = ( gb13["count"] < 10 )

gb13 = gb13[lc_mk].dropna().sort_values(by="mean", ascending=False).reset_index(drop=True)
px.bar(gb13[::-1], x="mean", y="brand", color="mean", color_continuous_scale=[px.colors.sequential.Plasma[0]] * 100)

Answer 1B)

Averaging by pure mean, and filtering for < 10 phones per brand, we uncover some "under the radar" smartphone brands. These include:

• Leitz (germany camera company)
• Nothing (minimalist new brand)
• Royale (only foldable screens)

gb12 = smarts.groupby(["brand"]).agg({"avg_rating": ["count", "mean"]}).reset_index()
gb12.columns = ["brand", "count", "mean"]
gb12.dropna().sort_values(by="mean", ascending=False).reset_index(drop=True)[:10]

	brand	count	mean
0	leitz	1	8.90
1	lenovo	2	8.80
2	sharp	1	8.80
3	asus	6	8.70
4	lg	2	8.70
5	royole	2	8.70
6	doogee	2	8.60
7	zte	2	8.55
8	nothing	5	8.52
9	blu	1	8.50

Answer 1B)

Averaging by pure mean, and filtering for >= 10 phones per brand, we uncover some "Consistency" brands in the local smartphone market. tThe best average rating is a tie between, beating out even Apple and Samsung etc, are:

• OnePlus
• Iqoo

# some companies make significantly more phones than others, and those are the largest brands
gb2 = smarts.groupby(["brand"]).agg({"avg_rating": ["count", "mean"]}).reset_index()
gb2.columns = ["brand", "count", "avg_rating_mean"]
gb2 = gb2.sort_values(by="avg_rating_mean", ascending=True)
g = px.bar(gb2[ct_mk].dropna(), y="brand", x="avg_rating_mean", color="avg_rating_mean", color_continuous_scale=[px.colors.sequential.Plasma[-1]] * 100)

g = px.scatter(gb2, x="avg_rating_mean", y="avg_rating_mean", hover_data="brand", size="count", color="count").update_traces({"marker":{"line": {"width": 1, "color": "black"}}}).update_layout({"width": 450, "height": 400})
g.add_annotation(x=7.77, y=8.1,
            text="Major smartphone <br>brands",
            showarrow=True,
            arrowhead=1)

gb14 = smarts.groupby(["brand"]).agg({"avg_rating": ["count", "mean"]}).dropna().reset_index()
gb14.columns = ["brand", "count", "mean"]

hc_mk = ( gb14["count"] >= 10 )

gb14 = gb14[hc_mk]

gb14.sort_values(by="mean", ascending=False).reset_index(drop=True)[:5]

	brand	count	mean
0	oneplus	38	8.223684
1	iqoo	32	8.221875
2	motorola	48	8.010417
3	huawei	11	8.000000
4	xiaomi	125	7.877600

data organized into a bar chart. The highest rated Consistency brands are visible quite close.

• This would be a valid answer to "best smartphone brand", as all brands are real smartphone brands (filtered for models >= 10)

px.bar(gb14.sort_values(by="mean", ascending=True), y="brand", x="mean", color="mean", color_continuous_scale=[px.colors.sequential.Plasma[-1]] * 100)

gb14.sort_values(by="mean", ascending=False)

	brand	count	mean
27	oneplus	38	8.223684
12	iqoo	32	8.221875
23	motorola	48	8.010417
9	huawei	11	8.000000
42	xiaomi	125	7.877600
28	oppo	82	7.873171
34	samsung	116	7.872414
30	poco	40	7.870000
8	honor	11	7.818182
7	google	12	7.733333
41	vivo	96	7.715625
0	apple	44	7.679545
31	realme	92	7.644565
11	infinix	28	7.628571
24	nokia	10	7.530000
38	tecno	31	7.367742

Part 1A)

organizing by a company's best phone (max)

gb3 = smarts.groupby(["brand"]).agg({"avg_rating": ["count", "max"]}).reset_index()
gb3.columns = ["brand", "count", "avg_rating_max"]
gb3 = gb3.sort_values(by="avg_rating_max", ascending=True)
gb3.sort_values(by="avg_rating_max", ascending=False)[:20].reset_index(drop=True)[:50]

	brand	count	avg_rating_max
0	realme	92	8.9
1	oppo	82	8.9
2	oneplus	38	8.9
3	motorola	48	8.9
4	lenovo	2	8.9
5	leitz	1	8.9
6	lg	2	8.9
7	samsung	116	8.9
8	tecno	31	8.9
9	sony	7	8.9
10	xiaomi	125	8.9
11	zte	2	8.9
12	google	12	8.9
13	huawei	11	8.9
14	infinix	28	8.9
15	iqoo	32	8.9
16	sharp	1	8.8
17	asus	6	8.8
18	nothing	5	8.8
19	honor	11	8.8

g = px.bar(gb3.dropna(), y="brand", x="avg_rating_max", color="avg_rating_max", color_continuous_scale="magma")
g.update_layout({"height": 400, "width":550, 
                 "yaxis": {"tickangle": 45, "tickfont": {"size": 8}}})

mx_mk =  ( gb3["avg_rating_max"] < 7 )
gb3[mx_mk]["brand"].unique()

array(['lyf', 'letv', 'tcl', 'itel', 'ikall', 'vertu', 'jio', 'leeco',
       'gionee', 'blackview', 'micromax'], dtype=object)

gb3.dropna().sort_values(by="avg_rating_max", ascending=False).reset_index(drop=True)

	brand	count	avg_rating_max
0	realme	92	8.9
1	oppo	82	8.9
2	oneplus	38	8.9
3	motorola	48	8.9
4	lenovo	2	8.9
5	leitz	1	8.9
6	lg	2	8.9
7	samsung	116	8.9
8	tecno	31	8.9
9	sony	7	8.9
10	xiaomi	125	8.9
11	zte	2	8.9
12	google	12	8.9
13	huawei	11	8.9
14	infinix	28	8.9
15	iqoo	32	8.9
16	sharp	1	8.8
17	asus	6	8.8
18	nothing	5	8.8
19	honor	11	8.8
20	vivo	96	8.8
21	doogee	2	8.8
22	nokia	10	8.7
23	royole	2	8.7
24	apple	44	8.6
25	nubia	6	8.6
26	poco	40	8.6
27	blu	1	8.5
28	lava	3	8.4
29	oukitel	3	8.4
30	tesla	1	8.3
31	redmi	3	8.1
32	cola	1	7.4
33	micromax	3	6.8
34	gionee	1	6.7
35	blackview	1	6.7
36	leeco	1	6.5
37	jio	1	6.4
38	vertu	1	6.2
39	itel	1	6.1
40	ikall	3	6.1
41	letv	1	6.0
42	tcl	1	6.0
43	lyf	1	6.0

Answer 1A)

there are 16 phone brands that have at least one phone with a maxed-out rating (8.9). these include: ['iqoo','infinix', 'huawei', 'google', 'realme', 'oppo', 'oneplus', 'motorola', 'lenovo', 'leitz', 'lg', 'samsung', 'tecno', 'sony', 'xiaomi', 'zte']

On the flip side, there are 10 brands whose top-rated phone has a rating of < 7.0. These include: 'lyf', 'letv', 'tcl', 'itel', 'ikall', 'vertu', 'jio', 'leeco', 'gionee', 'blackview', 'micromax'

insight: don't bet on any of the phone brands in the latter category

Step 7: What is the most popular model, among all brands?

There are 980 models. Which models scored the highest?

gb10 = smarts.groupby(["model", "brand"]).agg({"avg_rating": "mean"}).reset_index().sort_values(by="avg_rating", ascending=False)
g = px.bar(gb10[::-1].dropna(), y="model", x="avg_rating", color="avg_rating", color_continuous_scale="rainbow").update_traces(marker={"line": {"width": 0}, "opacity": 0.8})
g.add_shape(
    type="rect",
    x0=0.1, x1=9,
    y0=835, y1=880,
    line={"color": "white"}
).update_layout({"height": 500, "width": 1000})

# px.bar(gb10[:35][::-1].dropna(), x="model", y="avg_rating", color="avg_rating", color_continuous_scale="rainbow", range_color=[gb10["avg_rating"].min(), gb10["avg_rating"].max()]).update_traces(
#     marker={"line": {"width": 0}, "opacity": 0.8}
# ).update_layout({
#     "bargap": 0.1,
#     "height": 500, 
#     "yaxis": {"tickangle": 45, "tickfont": {"size": 9}}})

gb16 = smarts.groupby(["brand", "model"]).agg({"avg_rating": "mean"}).reset_index()
sm_mk = ( gb16["avg_rating"] == 8.9 )
pvt2 = gb16[sm_mk].pivot(columns="brand", index="model", values="avg_rating").T
px.imshow(pvt2, color_continuous_scale=[px.colors.sequential.Rainbow[-1]] * 100).update_layout({"height": 800, "width": 1600})

the answer is 34 way tie, from brands as above. Visual space between brands.

gb10[:40]

	model	brand	avg_rating
354	Oppo Find X6	oppo	8.9
635	Samsung Galaxy S23 Plus	samsung	8.9
133	Infinix Zero Ultra	infinix	8.9
665	Tecno Phantom X2 Pro	tecno	8.9
650	Sony Xperia 5 IV 5G	sony	8.9
312	OnePlus 9 Pro	oneplus	8.9
307	OnePlus 11 5G	oneplus	8.9
556	Samsung Galaxy A53 5G (8GB RAM + 128GB)	samsung	8.9
915	Xiaomi Redmi Note 12 Explorer	xiaomi	8.9
554	Samsung Galaxy A53 (8GB RAM + 256GB)	samsung	8.9
69	Google Pixel 6 Pro	google	8.9
96	Huawei Mate Xs 2	huawei	8.9
613	Samsung Galaxy S20 5G	samsung	8.9
607	Samsung Galaxy Note 10 Plus 5G	samsung	8.9
936	ZTE Axon 40 Ultra 5G	zte	8.9
211	Motorola ThinkPhone	motorola	8.9
606	Samsung Galaxy Note 10 Plus	samsung	8.9
940	iQOO 11 (16GB RAM + 256GB)	iqoo	8.9
164	Motorola Edge 30 Pro 5G	motorola	8.9
806	Xiaomi 12 Pro 5G	xiaomi	8.9
811	Xiaomi 12T Pro 5G	xiaomi	8.9
812	Xiaomi 13 5G	xiaomi	8.9
815	Xiaomi 14	xiaomi	8.9
150	Lenovo Legion Pro 2	lenovo	8.9
170	Motorola Edge Plus	motorola	8.9
641	Samsung Galaxy Z Fold 3	samsung	8.9
298	OnePlus 10 Pro 5G	oneplus	8.9
470	Realme GT 2 Pro 5G	realme	8.9
942	iQOO 11 Pro 5G	iqoo	8.9
208	Motorola Moto X40	motorola	8.9
471	Realme GT 2 Pro 5G (12GB RAM + 256GB)	realme	8.9
143	LG Wing 5G	lg	8.9
645	Sony Xperia 1 II	sony	8.9
169	Motorola Edge 40 Pro 5G	motorola	8.9
149	Leitz Phone 2	leitz	8.9
275	OPPO Find N Flip	oppo	8.8
744	Vivo X80 5G (12GB RAM + 256GB)	vivo	8.8
621	Samsung Galaxy S21 FE 5G (8GB RAM + 256GB)	samsung	8.8
619	Samsung Galaxy S21	samsung	8.8
359	Oppo Reno 10 Pro Plus	oppo	8.8

hr_mk = ( gb10["avg_rating"] == 8.9 )
len(gb10[hr_mk]) / len(gb10)

0.03571428571428571

Answer to question 2

There are 34 models that have the highet rating, as shown the table below. This represents 3.6% of all models.

Most major brands will have at least one top-rated phone.

gb10.reset_index(drop=True)[:50]

	model	brand	avg_rating
0	Oppo Find X6	oppo	8.9
1	Samsung Galaxy S23 Plus	samsung	8.9
2	Infinix Zero Ultra	infinix	8.9
3	Tecno Phantom X2 Pro	tecno	8.9
4	Sony Xperia 5 IV 5G	sony	8.9
5	OnePlus 9 Pro	oneplus	8.9
6	OnePlus 11 5G	oneplus	8.9
7	Samsung Galaxy A53 5G (8GB RAM + 128GB)	samsung	8.9
8	Xiaomi Redmi Note 12 Explorer	xiaomi	8.9
9	Samsung Galaxy A53 (8GB RAM + 256GB)	samsung	8.9
10	Google Pixel 6 Pro	google	8.9
11	Huawei Mate Xs 2	huawei	8.9
12	Samsung Galaxy S20 5G	samsung	8.9
13	Samsung Galaxy Note 10 Plus 5G	samsung	8.9
14	ZTE Axon 40 Ultra 5G	zte	8.9
15	Motorola ThinkPhone	motorola	8.9
16	Samsung Galaxy Note 10 Plus	samsung	8.9
17	iQOO 11 (16GB RAM + 256GB)	iqoo	8.9
18	Motorola Edge 30 Pro 5G	motorola	8.9
19	Xiaomi 12 Pro 5G	xiaomi	8.9
20	Xiaomi 12T Pro 5G	xiaomi	8.9
21	Xiaomi 13 5G	xiaomi	8.9
22	Xiaomi 14	xiaomi	8.9
23	Lenovo Legion Pro 2	lenovo	8.9
24	Motorola Edge Plus	motorola	8.9
25	Samsung Galaxy Z Fold 3	samsung	8.9
26	OnePlus 10 Pro 5G	oneplus	8.9
27	Realme GT 2 Pro 5G	realme	8.9
28	iQOO 11 Pro 5G	iqoo	8.9
29	Motorola Moto X40	motorola	8.9
30	Realme GT 2 Pro 5G (12GB RAM + 256GB)	realme	8.9
31	LG Wing 5G	lg	8.9
32	Sony Xperia 1 II	sony	8.9
33	Motorola Edge 40 Pro 5G	motorola	8.9
34	Leitz Phone 2	leitz	8.9
35	OPPO Find N Flip	oppo	8.8
36	Vivo X80 5G (12GB RAM + 256GB)	vivo	8.8
37	Samsung Galaxy S21 FE 5G (8GB RAM + 256GB)	samsung	8.8
38	Samsung Galaxy S21	samsung	8.8
39	Oppo Reno 10 Pro Plus	oppo	8.8
40	iQOO 9 Pro 5G	iqoo	8.8
41	Sharp Aquos R5G	sharp	8.8
42	Vivo X90 5G	vivo	8.8
43	Xiaomi 11T Pro 5G (12GB RAM + 256GB)	xiaomi	8.8
44	Samsung Galaxy S21 Plus	samsung	8.8
45	Asus ROG Phone 6 Batman Edition	asus	8.8
46	Honor 70 5G	honor	8.8
47	Samsung Galaxy S20 FE 5G	samsung	8.8
48	Samsung Galaxy S20	samsung	8.8
49	Samsung Galaxy S23	samsung	8.8

Step 8: what is the most expensive smartphone model?

Many people consider a phone a good investment. Let's see what happens when people consider the price as part of the value.

gb5 = smarts.groupby(["model", "brand"]).agg({"price": "mean"}).reset_index()
gb5 = gb5.sort_values(by="price", ascending=False)
gb5[:10]

	model	brand	price
689	Vertu Signature Touch	vertu	650000.0
866	Xiaomi Redmi K20 Pro Signature Edition	xiaomi	480000.0
94	Huawei Mate 50 RS Porsche Design	huawei	239999.0
91	Huawei Mate 30 RS Porsche Design	huawei	214990.0
830	Xiaomi Mi Mix Alpha	xiaomi	199990.0
32	Apple iPhone 14 Pro Max (1TB)	apple	182999.0
19	Apple iPhone 13 Pro Max (1TB)	apple	179900.0
29	Apple iPhone 14 Pro (1TB)	apple	172999.0
34	Apple iPhone 14 Pro Max (512GB)	apple	169900.0
95	Huawei Mate X	huawei	169000.0

Answer:

the "Vertu Signature Touch" is the most expensive model. See below visualization for just how expensive.

g = px.scatter(gb5[::-1], y="model", x="price", color="price", color_continuous_scale="purp")

g.add_annotation(
    text="iPhone 14 Pro Max",
    x=185000, 
    y=960,
    ax=196000,
    ay=850,
    showarrow=True,
    arrowhead=1,
    xref='x',
    yref='y',
    axref='x',
    ayref='y',    
)
g.update_layout({"height": 500, "width": 1600,
                 "yaxis": {"tickangle": 45, "tickfont": {"size": 9}}}).update_traces(marker={"line": {"width": 0}, "opacity": 0.8}).update_traces({"marker":{"line": {"width": 0.3, "color": "purple"}}})

smarts[smarts["model"]=="Vertu Signature Touch"]

	brand	model	price	avg_rating	processor_speed	battery_capacity	primary_camera_rear	total_res
732	vertu	Vertu Signature Touch	650000	6.2	1.5	2275.0	13.0	2073600

g = px.scatter(smarts, y="avg_rating", x="price", color="price", color_continuous_scale=px.colors.sequential.Purp[::-1]).update_traces({"marker":{"line": {"width": .5, "color": "purple"}}}).update_layout({"width": 550})

g.add_annotation(
    x=640000, 
    y=6.4,
    showarrow=True,
    arrowhead=1,
    text="The Vertu Signature Touch"
)

Step 8: What is the fastest phone?

Units range from 1.20 - 3.22.

gb6 = smarts.groupby(["model", "brand"]).agg({"processor_speed": "mean"}).sort_values(by="processor_speed", ascending=False).reset_index()

gb6[:20]

	model	brand	processor_speed
0	Apple iPhone 14	apple	3.22
1	Apple iPhone 13 Pro Max (256GB)	apple	3.22
2	Apple iPhone 14 (256GB)	apple	3.22
3	Apple iPhone 14 (512GB)	apple	3.22
4	Apple iPhone 13 Pro (256GB)	apple	3.22
5	Apple iPhone 13 Pro Max	apple	3.22
6	Apple iPhone 13 Pro (1TB)	apple	3.22
7	Apple iPhone 14 Plus (256GB)	apple	3.22
8	Apple iPhone 13 Pro Max (1TB)	apple	3.22
9	Apple iPhone 14 Plus (512GB)	apple	3.22
10	Apple iPhone 14 Plus	apple	3.22
11	Apple iPhone SE 3 2022	apple	3.22
12	Apple iPhone 13	apple	3.22
13	Apple iPhone 13 Mini	apple	3.22
14	Apple iPhone SE 4	apple	3.22
15	Apple iPhone 13 Pro	apple	3.22
16	Apple iPhone 13 (512GB)	apple	3.22
17	Apple iPhone 13 (256GB)	apple	3.22
18	OnePlus 10T (12GB RAM + 256GB)	oneplus	3.20
19	iQOO Neo 7 Racing Edition	iqoo	3.20

Answer:

The top processing speed is a 15-way tie at 3.22 - all of Apple products.

However, there are 96 other phones with just a slightly slower processor at 3.20, from companies like OnePlus, Xiaomi, Iqoo, and others.

g = px.bar(gb6.dropna()[::-1], y="model", x="processor_speed", color="processor_speed", color_continuous_scale=px.colors.sequential.YlOrRd_r).update_traces(marker={"line": {"width": 0}, "opacity": 1})

g.add_shape(
    type="rect",
    x0=0.05, x1=3.23,
    y0=915, y1=945,
    line={"color": "red"}
).update_layout({"height": 500, "width": 1000})

g = px.bar(gb6.dropna()[::-1], y="model", x="processor_speed", color="processor_speed", color_continuous_scale=px.colors.sequential.YlOrRd_r).update_traces(marker={"line": {"width": 0}, "opacity": 1})
g.add_shape(
    type="rect",
    x0=0.1, x1=3.21,
    y0=810, y1=925,
    line={"color": "#6d84c8"}
).update_layout({"height": 500, "width": 1000})

fn_mk = ( gb6["processor_speed"] >= 3.2 )
gb6.dropna()[fn_mk][::-1]["processor_speed"].unique(),

(array([3.2 , 3.22]),)

px.bar(gb6[fn_mk][::-1], y="model", x="processor_speed")

trace0 = go.Bar(
    x=gb6.dropna()[fn_mk][::-1]["processor_speed"].unique(),
    y=["top performing @ 3.22", "top performing1 @ 3.20"],
    orientation="h",
    marker={"color": px.colors.sequential.YlOrRd_r[6]},
    name="top performing"
)
trace1 = go.Bar(
    x=gb6.dropna()[::-1][:2]["processor_speed"],
    y=gb6.dropna()[::-1][:2]["model"],
    orientation="h",
    marker={"color": px.colors.sequential.YlOrRd_r[3]},
    name="bottom performing"
)

go.Figure([trace1, trace0]).update_layout({"height": 600})

as shown here, the top 2 categories are close enough not to really make a substantial difference. There are 94 models at 3.20, in addition to the 17 at 3.22, which means that 111 models have a very fast processing speed.

• 15 models == 3.22, all Apple
• 94 models == 3.20, negligeably different (0.667% percent difference)
  • -> 11.1% of phone models have speeds 3.20 or faster
• These brands are:
  • 'oneplus', 'iqoo', 'xiaomi', 'asus', 'oppo', 'vivo', 'realme', 'poco', 'nubia', 'nothing', 'nokia', 'samsung', 'motorola', 'huawei', 'zte'
  • Not just Apple with the fastest phones - even if they'd like to say so!

pr_mk = ( gb6["processor_speed"] == 3.20 )

len(gb6[pr_mk])

94

The following brands have processng speeds in that 3.20 category bucket.

len(gb6[pr_mk]["brand"].unique()), gb6[pr_mk]["brand"].unique()

(15,
 array(['oneplus', 'iqoo', 'xiaomi', 'asus', 'oppo', 'vivo', 'realme',
        'poco', 'nubia', 'nothing', 'nokia', 'samsung', 'motorola',
        'huawei', 'zte'], dtype=object))

Step 9: what phone has the best camera?

gb7 = smarts.groupby(["model", "brand"]).agg({"primary_camera_rear": "mean"}).sort_values(by="primary_camera_rear", ascending=False).reset_index()
g = px.bar(gb7[::-1], y="model", x="primary_camera_rear", color="primary_camera_rear", color_continuous_scale="purpor").update_traces(marker={"line": {"width": 0}, "opacity": 0.2})
g.update_layout({"height": 600, "width": 1600})

hc_mk = ( gb7["primary_camera_rear"] == 200 )
gb7[hc_mk]["brand"].unique()

array(['nokia', 'xiaomi', 'samsung', 'motorola', 'infinix', 'honor'],
      dtype=object)

Answer

A 17-way tie among different models with 200 megapixels. AFter that, there is a steep drop off to 108 megapixels.

Brands with the best camera include:

• 'nokia',
• 'xiaomi',
• 'samsung',
• 'motorola',
• 'infinix',
• 'honor'

and notably, not Apple.

gb7[:20]

	model	brand	primary_camera_rear
0	Nokia N73 5G	nokia	200.0
1	Nokia X60 5G	nokia	200.0
2	Xiaomi Redmi Note 12 Discovery Edition	xiaomi	200.0
3	Xiaomi Redmi Note 12 Explorer	xiaomi	200.0
4	Samsung Galaxy S24 Ultra	samsung	200.0
5	Nokia X60 Pro 5G	nokia	200.0
6	Motorola Edge 30 Ultra	motorola	200.0
7	Motorola Edge 30 Ultra (12GB RAM + 256GB)	motorola	200.0
8	Motorola Moto X30 Pro	motorola	200.0
9	Motorola Moto Edge X30 Pro	motorola	200.0
10	Samsung Galaxy S23 Ultra 5G	samsung	200.0
11	Infinix Zero Ultra	infinix	200.0
12	Xiaomi Redmi Note 12 Pro Plus	xiaomi	200.0
13	Xiaomi Redmi Note 12 Pro Plus (12GB RAM + 256GB)	xiaomi	200.0
14	Xiaomi Redmi Note 12 Pro Plus (6GB RAM + 128GB)	xiaomi	200.0
15	Honor 80 Pro Plus	honor	200.0
16	Xiaomi 12T Pro 5G	xiaomi	200.0
17	Samsung Galaxy Note 30 Ultra 5G	samsung	200.0
18	OnePlus Nord CE 3 5G	oneplus	108.0
19	Xiaomi Redmi Note 10 Pro Max (6GB RAM + 128GB)	xiaomi	108.0

Step 10: What phone model has the best resolution?

*this is an engineered feature, calculated by multiplying pixels height x pixels width.

gb8 = smarts.groupby(["model", "brand"]).agg({"total_res": "mean"}).reset_index().sort_values(by="total_res", ascending=False)
g = px.bar(gb8[::-1], y="model", x="total_res", color="total_res", color_continuous_scale="blues").update_traces(marker={"line": {"width": 0}, "opacity": 0.8})
g.update_layout({"height": 600, "width": 1600})

gb8[:5]

	model	brand	total_res
646	Sony Xperia 1 IV (12GB RAM + 512GB)	sony	6312960.0
645	Sony Xperia 1 II	sony	6312960.0
653	Sony Xperia Pro-I	sony	6312960.0
95	Huawei Mate X	huawei	5456000.0
96	Huawei Mate Xs 2	huawei	5456000.0

Answer

Sony owns all models with the highest resolution. The model of phone is the Sony Xperia.

This is another good example of smartphone brands specializing.

idx = smart[smart["brand"]=="sony"].filter(regex="\w*[Xperia]w\*", axis=1).index
smarts.loc[idx, ::].sort_values(by="total_res", ascending=False)

	brand	model	price	avg_rating	processor_speed	battery_capacity	primary_camera_rear	total_res
688	sony	Sony Xperia 1 II	69999	8.9	2.84	4000.0	12.0	6312960
689	sony	Sony Xperia 1 IV (12GB RAM + 512GB)	92980	NaN	3.00	5000.0	12.0	6312960
696	sony	Sony Xperia Pro-I	134999	NaN	2.84	4500.0	12.0	6312960
691	sony	Sony Xperia 10 III Lite 5G	30990	7.8	2.00	4500.0	12.0	2721600
690	sony	Sony Xperia 10 II	27999	7.5	2.00	3600.0	12.0	2721600
692	sony	Sony Xperia 5 II	69990	8.6	2.84	4000.0	12.0	2721600
693	sony	Sony Xperia 5 IV 5G	82199	8.9	3.00	5000.0	12.0	2721600
694	sony	Sony Xperia Ace IV	19990	8.0	2.00	5000.0	50.0	2592000
695	sony	Sony Xperia L5 5G	15990	7.3	2.30	4000.0	13.0	1209600

Step 11: Best battery capacity

gb9 = smarts.groupby(["model", "brand"]).agg({"battery_capacity": "mean"}).reset_index().sort_values(by="battery_capacity", ascending=False)
g = px.bar(gb9[::-1], y="model", x="battery_capacity", color="battery_capacity", color_continuous_scale="Emrld").update_traces(marker={"line": {"width": 0}, "opacity": 0.8})
g.update_layout({"height": 600, "width": 1600})

gb9[:5]

	model	brand	battery_capacity
58	Doogee V Max	doogee	22000.0
366	Oukitel WP19	oukitel	21000.0
367	Oukitel WP21	oukitel	9800.0
368	Oukitel WP9	oukitel	8000.0
674	Tecno Pova Neo 2	tecno	7000.0

Answer

The Doogee V Max has the best battery capacity, with the Ouitel WP19 coming in at a close second.

• The next largest battery is less tha 50% of these top two models.
• It appears that the brand Doogee specializes in long battery life.

Conclusions

1. Apple not market leader by many measurements
   • some specs
     • battery life
     • camera
     • resolution
   • as well as ratings
   • yet is still one of the most expensive
2. Other "Top Tier" brands very much so present and popular
   • Samsung
   • Haiwei
   • Xiaomi
3. "Consistency" brands are mostly local
   • these are the brands that consistently produce highly rated phones in the market (highest pure mean of all phones)
   • OnePlus leads in this measurement
   • 4/5 of "Consistency" brands are local to Asia
     • OnePlus, Iqoo, Xiaomi, Motorola, Huawei
4. Experimental forays
   • Leitz: Germany camera company; Tesla: car company
   • Points to the competitive nature of the market
5. Specialization
   • for some of the specs, there is one brand that is the best at that spec
     • battery: Doogee
     • resolution: Sony
     • foldable: Royole
6. "Niche" branding
   • sometimes as differentiating as a spec, sometimes combined
     • "high end luxury" + price (+ real time assistant avialable): Vertu
     • "rugged" + best battery life: Doogee
     • "minimalism": Nothing phone

###################################################################################################################################################

Bonus data vis: How much are people paying for their processors? Color as processor speed.

px.scatter_3d(smart, x="brand", y="processor_speed", z="price", color="processor_speed", symbol_sequence=["square"]).update_layout({"height":500})

pr_mk = ( smart["processor_speed"] >= 3.2)
na_mk = ( smart["brand"] != "apple" ) 
smarts[pr_mk & na_mk].sort_values(by="price", ascending=True)

	brand	model	price	avg_rating	processor_speed	battery_capacity	primary_camera_rear	total_res
978	zte	ZTE Axon 30S	19999	8.2	3.2	4200.0	50.0	2656800
214	motorola	Motorola Edge S	22490	8.7	3.2	5000.0	64.0	2721600
152	iqoo	iQOO Neo 6 SE 5G	22990	8.3	3.2	4700.0	64.0	2592000
869	xiaomi	Xiaomi Mi 11X	23890	8.0	3.2	4520.0	48.0	2592000
141	iqoo	iQOO 7	24990	8.3	3.2	4400.0	48.0	2592000
748	vivo	Vivo T2 5G	24990	8.5	3.2	4700.0	64.0	2592000
519	realme	Realme GT Neo 3T	24999	7.9	3.2	5000.0	64.0	2592000
870	xiaomi	Xiaomi Mi 11X 5G (8GB RAM + 128GB)	25990	8.3	3.2	4520.0	48.0	2592000
419	poco	Poco F4 5G	25999	8.2	3.2	4500.0	64.0	2592000
520	realme	Realme GT Neo 3T (8GB RAM + 128GB)	26499	8.1	3.2	5000.0	64.0	2592000
420	poco	Poco F4 5G (8GB RAM + 128GB)	27999	8.5	3.2	4500.0	64.0	2592000
521	realme	Realme GT Neo 3T (8GB RAM + 256GB)	28499	8.2	3.2	5000.0	64.0	2592000
150	iqoo	iQOO Neo 6 5G	28999	8.2	3.2	4700.0	64.0	2592000
450	poco	POCO X5 GT	29990	8.4	3.2	5000.0	64.0	2592000
734	vivo	Vivo S16	29990	8.5	3.2	4600.0	64.0	2592000
914	xiaomi	Xiaomi Redmi K60	29990	8.4	3.2	5500.0	64.0	4608000
424	poco	Poco F5 Pro	29999	8.4	3.2	5500.0	64.0	4608000
418	poco	Poco F4 (12GB RAM + 256GB)	29999	8.6	3.2	4500.0	64.0	2592000
524	realme	Realme GT Neo2 5G	31999	8.2	3.2	5000.0	64.0	2592000
154	iqoo	iQOO Neo 7 Racing Edition	32990	8.4	3.2	5000.0	50.0	2592000
516	realme	Realme GT Explorer Master Edition	32990	8.5	3.2	4500.0	50.0	2592000
151	iqoo	iQOO Neo 6 5G (12GB RAM + 256GB)	33749	8.4	3.2	4700.0	64.0	2592000
422	poco	Poco F5	34990	8.5	3.2	4700.0	64.0	2592000
278	nubia	Nubia Z50	34999	8.2	3.2	5000.0	64.0	2592000
296	oneplus	OnePlus Ace 2	34999	8.4	3.2	5000.0	50.0	2604960
523	realme	Realme GT Neo 5	34999	8.3	3.2	4600.0	50.0	2604960
142	iqoo	iQOO 7 (12GB RAM + 256GB)	35990	8.5	3.2	4400.0	48.0	2592000
916	xiaomi	Xiaomi Redmi K60 Pro	38999	8.5	3.2	5000.0	54.0	4608000
266	nokia	Nokia X60 5G	39990	8.7	3.2	6000.0	200.0	2592000
608	samsung	Samsung Galaxy A83 5G	39990	8.6	3.2	4500.0	64.0	4608000
272	nothing	Nothing Phone 2	39990	8.8	3.2	4700.0	50.0	2592000
290	oneplus	OnePlus 11R	39999	8.5	3.2	5000.0	50.0	2604960
911	xiaomi	Xiaomi Redmi K50 Ultra 5G	39999	8.5	3.2	5000.0	108.0	3308640
251	motorola	Motorola Moto X40	39999	8.9	3.2	4600.0	50.0	2592000
512	realme	Realme GT 2 Explorer Master Edition	40990	8.6	3.2	5000.0	50.0	2604960
250	motorola	Motorola Moto X30 Pro	41990	NaN	3.2	4450.0	200.0	2592000
211	motorola	Motorola Edge 40 Fusion	42990	8.7	3.2	4500.0	50.0	2592000
776	vivo	Vivo V26 Pro	42990	8.7	3.2	4800.0	64.0	2592000
778	vivo	Vivo V27 Pro	43999	8.7	3.2	5000.0	108.0	2592000
392	oppo	OPPO Reno 6 Pro Plus 5G	44990	8.7	3.2	4500.0	50.0	2592000
285	oneplus	OnePlus 10T	44999	8.5	3.2	4800.0	50.0	2604960
402	oppo	OPPO Reno 9 Pro Plus	45999	8.6	3.2	4700.0	50.0	2604960
275	nubia	Nubia Red Magic 8 Pro 5G	46990	8.4	3.2	6000.0	50.0	2767680
855	xiaomi	Xiaomi 13 5G	46990	8.9	3.2	4500.0	50.0	2592000
149	iqoo	iQOO 9T 5G	47990	8.5	3.2	4700.0	50.0	2592000
298	oneplus	OnePlus Ace Pro	47990	8.6	3.2	4800.0	50.0	2604960
267	nokia	Nokia X60 Pro 5G	49990	NaN	3.2	6000.0	200.0	2600640
254	motorola	Motorola ThinkPhone	49990	8.9	3.2	5000.0	50.0	2592000
286	oneplus	OnePlus 10T (12GB RAM + 256GB)	50999	8.7	3.2	4800.0	50.0	2604960
915	xiaomi	Xiaomi Redmi K60 Gaming Edition	54990	8.6	3.2	5000.0	50.0	4608000
288	oneplus	OnePlus 11 5G	54999	8.9	3.2	5000.0	50.0	4631040
858	xiaomi	Xiaomi 14	54999	8.9	3.2	5000.0	50.0	4608000
208	motorola	Motorola Edge 30 Ultra	54999	NaN	3.2	4610.0	200.0	2592000
287	oneplus	OnePlus 10T (16GB RAM + 256GB)	55999	8.7	3.2	4800.0	50.0	2604960
856	xiaomi	Xiaomi 13 Pro 5G	58990	NaN	3.2	4820.0	50.3	4608000
386	oppo	Oppo Reno 10 Pro Plus	58990	8.8	3.2	4700.0	64.0	2604960
140	iqoo	iQOO 11 Pro 5G	58990	8.9	3.2	4700.0	50.0	4608000
854	xiaomi	Xiaomi 12T Pro 5G	59990	8.9	3.2	5000.0	200.0	3308640
212	motorola	Motorola Edge 40 Pro 5G	59990	8.9	3.2	4800.0	50.0	2656800
224	motorola	Motorola Moto Edge X30 Pro	59990	NaN	3.2	4450.0	200.0	2527200
209	motorola	Motorola Edge 30 Ultra (12GB RAM + 256GB)	59999	NaN	3.2	4610.0	200.0	2592000
139	iqoo	iQOO 11 5G	59999	8.8	3.2	5000.0	50.0	4608000
274	nubia	Nubia Red Magic 7S Pro	60990	8.6	3.2	5000.0	64.0	2592000
276	nubia	Nubia Red Magic 8 Pro Plus	60999	8.6	3.2	5000.0	50.0	2767680
52	asus	Asus ZenFone 9	63990	8.6	3.2	4300.0	50.0	2592000
138	iqoo	iQOO 11 (16GB RAM + 256GB)	64999	8.9	3.2	5000.0	50.0	4608000
376	oppo	Oppo Find X6	69990	8.9	3.2	4700.0	50.0	2592000
852	xiaomi	Xiaomi 12S Ultra	69990	NaN	3.2	4860.0	50.0	4608000
291	oneplus	OnePlus 12 Pro	69999	NaN	3.2	4800.0	108.0	4913280
289	oneplus	OnePlus 11 Pro	69999	NaN	3.2	4500.0	50.0	4631040
676	samsung	Samsung Galaxy S23	70990	8.8	3.2	3900.0	50.0	2527200
373	oppo	OPPO Find N2 Flip	70990	8.8	3.2	4300.0	50.0	2721600
47	asus	Asus ROG Phone 6	71999	8.6	3.2	6000.0	50.0	2643840
857	xiaomi	Xiaomi 13 Ultra 5G	71999	NaN	3.2	5000.0	108.0	4608000
48	asus	Asus ROG Phone 6 Batman Edition	72999	8.8	3.2	6000.0	50.0	2643840
792	vivo	Vivo X90 Pro Plus 5G	73999	NaN	3.2	4700.0	50.3	4608000
51	asus	Asus ROG Phone 7	75990	8.7	3.2	6000.0	64.0	2643840
851	xiaomi	Xiaomi 12 Ultra 5G	75999	NaN	3.2	4800.0	50.0	4608000
377	oppo	Oppo Find X6 Pro	77990	NaN	3.2	4700.0	50.0	2592000
789	vivo	Vivo X80 Pro Plus 5G	82990	NaN	3.2	4500.0	50.0	4608000
678	samsung	Samsung Galaxy S23 Plus	84990	8.9	3.2	4700.0	50.0	2527200
49	asus	Asus ROG Phone 6 Pro 5G	89999	8.8	3.2	6000.0	50.0	2643840
683	samsung	Samsung Galaxy Z Flip 4 5G	89999	8.8	3.2	3700.0	12.0	2851200
372	oppo	Oppo Find N2 5G	94990	NaN	3.2	4520.0	50.0	3440640
371	oppo	Oppo Find N Fold	99990	NaN	3.2	5000.0	50.0	3440640
652	samsung	Samsung Galaxy Note 30 Ultra 5G	104999	NaN	3.2	4500.0	200.0	4628160
875	xiaomi	Xiaomi Mix Fold 2 5G	106990	NaN	3.2	4500.0	50.0	4134240
50	asus	Asus ROG Phone 6D Ultimate	107990	NaN	3.2	6000.0	50.0	2643840
784	vivo	Vivo X Fold Plus	113990	NaN	3.2	4730.0	50.0	4138560
679	samsung	Samsung Galaxy S23 Ultra 5G	114990	NaN	3.2	5000.0	200.0	4446720
781	vivo	Vivo X Fold 2	119990	NaN	3.2	4800.0	50.0	4138560
685	samsung	Samsung Galaxy Z Fold 4	154998	NaN	3.2	4400.0	50.0	3942912
686	samsung	Samsung Galaxy Z Fold 4 (12GB RAM + 1TB)	163980	NaN	3.2	4400.0	50.0	3942912
94	huawei	Huawei Mate 50 RS Porsche Design	239999	8.1	3.2	4700.0	50.0	3170592